Wage stickiness for incumbents vs. new workers

Masao Fukui, job market candidate from MIT, has made some significant progress on this problem, paper here.  You should cringe if you just hear ‘wage stickiness” — for the incumbents, maybe, due to morale effects, because a grumpy worker who just took a pay cut might wreck things.  But why is there wage stickiness for the new, not yet hired workers?  Isn’t the new wage bargain what they need to negotiate in the first place?  Other than postulating stubborn unemployed workers who overestimate their worth, how might we generate microfoundations for wage stickiness for the not yet hired, also known as “the unemployed”?  Here is Fukui’s abstract:

I develop a new theory of wage rigidity and unemployment fluctuations. The starting point of my analysis is a generalized version of Burdett and Mortensen’s (1998) job ladder model featuring risk-neutral firms, risk-averse workers, and aggregate risk. Because of on-the-job search, my model generates wage rigidity both for incumbent workers, through standard insurance motives, and for new hires, through novel strategic complementarities in wage setting between firms. In contrast to the conventional wisdom in the macro literature, the introduction of on-the-job search implies that: (i) the wage rigidity of incumbent workers, rather than new hires, is the critical determinant of unemployment fluctuations; (ii) fairness considerations in wage setting dampen, rather than amplify, unemployment fluctuations; and (iii) new hire wages are too flexible, rather than too rigid, in the decentralized equilibrium. Quantitatively, the wage rigidity of incumbent workers caused by the insurance motive alone accounts for about one fifth of the unemployment fluctuations observed in the data.

As for wage stickiness for the not yet hired workers, here is I think the key point:

I show using simple phase diagrams that new hire wages must always feature rigidity at the top of the job ladder. This comes from the fact that at the very top of the job ladder, potential new employers have no incentive to increase wages above what the incumbent firms offer because there would be no additional workers to poach. This extremely strong strategic complementarity spills over toward lower job ladder rungs, and the wages are asymptotically rigid regardless of functional forms or parameter values. This result provides an explanation for the recent evidence on new hire wage rigidity.

The paper has many other interesting features. For instance, once wage rigidity for incumbent workers is a larger cause of unemployment, as opposed to just wage rigidity for new hires, the Shimer empirical critiques of labor market matching models dissipate.  So matching models are strengthened, as are models of real rather than nominal rigidity of wages.

I am not yet sure if Fukui is right, but in any case this paper is a major contribution to the theory of wage-setting and it seems he is getting closer to the truth than anyone else has.

Tricky stuff!  Via Ivan Werning.